Berson Furan Maleic Anhydride 1953 Ja01103a060
Transcript of Berson Furan Maleic Anhydride 1953 Ja01103a060
8/10/2019 Berson Furan Maleic Anhydride 1953 Ja01103a060
http://slidepdf.com/reader/full/berson-furan-maleic-anhydride-1953-ja01103a060 1/6
April 5
1953
THE
TEREOCHEMISTRY O F THE FURAN-MALEIC
CIDREACTION
1721
then separated, washed with water and
10
sodium bicar-
bonate (both precooled in ice-bath), dried over sodium sul-
fa te and the solvent removed. The final traces of benzene
were removed by azeotropic distillation with ethanol and
the product crystallized directly from this latter solvent;
yield 245 mg. (82%), m.p. 128-129', [ a I z 6 D +41.5'. The
cholestanone has an infrared spectrum identical with au-
thentic material an d displayed no bands a t the C-D stretch-
ing frequency. Deuterium analysis showed only the nat-
ural at om per cent. excess D.
Decomposition of Reduction Complex with DzO.-The re-
duction was conducted as described above using 0.9 g. of
cholestanone enol acetate and 4.2 mmoles of lith ium alumi-
num hydride. After the solution had been stirred for one
hour, 0.5
g.
(25 mmoles) of DzO (99.98%) was added and
the solution allowed to stand under
a
nitrogen atmosphere
for 10 hours. Additional anhydrous ether was added, the
mixture stirred for another two hours and then centrifuged.
The ether was decanted and the residue was leached twice
with additional anhydrous ether. Evaporation of th e fil-
tered decantate yielded 685 mg. of solid which was dis-
solved in hexane and chromatographed on alumina which
had been neutralized and activated.Ia After elution of the
cholestanone (251 mg.), with 15% ether in hexane, a volume
of this same solvent containing
a
few per cent. methanol
was passed through the column to elute the sterols (412
mg.).l4 A second chromatograph on less active aluminal6
yielded 100 mg. of epicholestanol-2-d, m.p. 180-182', and
286 mg. of cholestanol-2-d, m . t 140-142', recrystallized,
m.p. 141.5-142.5', [aIz6D +23 . Both of these steroids
show the characteristic C-D stretching frequency. Deu-
terium analysis of the cholestanols showed 2.20 atom per
cent. excess D (theory 2.08 atom per cent. excess D) .
A portion of the cholestanol-2-d was oxidized to choles-
tanone-2-d in the same manner as described for the oxid:-
tion of cholestanol-3-d. The product, m.p. 129.5-130.0
,
retained t he C-D absorption band
at
2060 cm.-I and deu-
(13)
C.
W. Shoppee and G. H. R. Summers, J . Chcm. Soc., 687
(1950).
14)
Th e alumina employed for th is chromatograph was sufficiently
active to prevent r eady elution of hydroxylated material by th e ordi-
nary solvents, ether and hexane. While the sterols were absorbed
on this very active alumina, the solution containing methanol was
passed through the column in order to wash
off
an y deuterium from
th e oxygen of the alcohol. Elution in this manner did no t give a
satisfacto ry separation of th e sterols
so
it was necessary
to
perform a
second chromatograph on less active alumina in order t o separate th e
epimers.
(15) Merck, Aluminum Oxide Reagent, suitable for chromatographic
adsorption.
terium analysis shows a 2.28 atom per cent. excess D.
A
sample of this ketone was dissolved in hexane and passed
onto a column of basic alumina.16
After a quantity of hex-
ane saturated with water had been passed through the
column, the cholestanone was eluted with
15%
ether in
hexane. This ketone, as directly eluted, had
a
m.p. 126-
128' and
no
longer showed any absorption in the C-D
stretching frequency region. Deuterium analysis showed
only the natural atom per cent. excess
D.
Reduction of Cholestenone Enol Acetate with
Varying
Amounts
of
Lithium Aluminum Hydride.-The experiments
were performed a t 25' in a manner described previously for
an inverse reduction.* The only significant difference was
that saturated sodium potassium tartrate solution was em-
ployed to decompose the reduction complex in place of di-
lute mineral acid. Considerable attention was given to the
maintepance of anhydrous conditions, to th e employment
of scrupulously dried equipment and t o th e standardization
of the hydride solutions used. The stock hydride solution
was prepared, stored and standardized both by titration
and hydrogen evolution before each
run .
All equipment
was dried a t 100' in a vacuum oven immediately prior to
use and the cholestenone enol acetate dried at room tem-
perature for 10 hours at
10 6
mm. All experiments were
performed with 0.500 g. of ester in
10
ml. of anhydrous
ether . The reaction products were separated by chroma-
tography in the usual manner on alumina. The results
are listed in Table I.
Attempted Reduction
of
Cholestenone Enolate Ion.-To
50 ml. of dry butyl ether there was added, in turn , 0.45 ml.
(4.4 mmoles) of freshly distilled diethylamine, 3.3 mmoles
of phenyllithium in 4.6 ml. ethereal solution and after three
hours, 0.80 g. (2.1 mmoles) of cholestenone. The mixture
was stirred for four hours at room temperature in a nitrogen
atmosphere and was then heated on a steam-bath and ap-
proximately
10
ml. of solvent was removed under reduced
pressure.
A solution
of 5
mmoles of l ithium aluminum hydride
in 3.5 ml. of ethyl ether was added to the butyl ether solu-
tion and t he mixture stirred overnight at room temperature.
After decomposition of the reaction complex, the mixture
was processed in the standard fashion. There was obtained,
248 mg. (34%) of cholestenone, m.p. 76-78'; 325 mg.
(45%)
of
mixed A4-cholesten-3-ols, m.p. 120S1380; 127 mg.
(17%) of A4-cholesten-3P-ol, m.p. 131-133
.
(16) Fisher, Alumina Adsorption.
This alumina had previously
been shown to hav e sufficient basic streng th
to
cause the hydrolysis of
cholesteryl acetate during normal chromatographic procedures.
BERKELEY, CALIFORNIA
[CONTRIBUTIONROM THE DEPARTMENTF CHEMISTRY,NIVERSITY
F
SOUTHERNALIFORNIA
Th e Stereochemistry of the Furan-Maleic Acid Reac tion1
BY
JEROME A.
BERSON ND RONALD
WIDLER
RECEIVED OVEMBER, 1952
The reaction of furan with maleic acid in. water is shown to be stereochemically heterogeneous.
endo-Addition is kinetic-
ally favored bu t the endo-adduct gradually isomerizes to the more stable exo-adduct.
A
method is described for analysis of
the diene reaction mixtures. The free energy change for isomerization of
endo
to exo-adduct is estimated to be n ot more
than
-1.2
kcal./mole. A discussion is given of the possible origins of energy differences in endo- and exo-isomers of 2,2,1-
bicycloheptane derivatives.
The general occurrence of endo-stereospecificity
in the products of diene additions investigated in a
series of early studies was summarized in the well-
known Alder rule.2 Fur ther investgations,a-s
stimulated by the brief report2 that the diene
additions of fulvenes were stereochemically non-
(1)
Thi s work reported here will form a portion of t he thesis to be
submitte d by Ron ald Swidler in part ial fulfillment of t he requiremen ts
for the degree of Doctor of Philosophy.
(2)
K.
Adler and 0 tein,
Angc w .
Chem.,
50 ,
614
(1937).
(3) R.
B.Woodward atid
H.
Baer, THIS OURNAL 66 645 1944).
4) K. Alder, F. W. Chamber sand W.Trimborn, Ann
566 27
1950).
5 ) K
Alder
and R R uhma nn, i b i d .
566,
\10.10).
homogeneous, brought to light the dependence of
the steric outcome
of
these additions upon the
reaction conditions, the endo-adduct being formed
more rapidly but giving way a t longer reaction
times or higher temperatures to the thermo-
dynamically-favored exo-isomer. The reaction of
furan with maleimidea now seems to conform to this
energetic scheme, and even in the difficultly mobile
eyclopentadiene-maleic anhydride endo-exo adduct
equilibrium,' the same kinetic order and relative
(6) H Kwart and I. Burchuk, THIS OURNAL 7 4 ,
3094 1952) .
7 ) D Craig,
r btd
73
990
10.71)
8/10/2019 Berson Furan Maleic Anhydride 1953 Ja01103a060
http://slidepdf.com/reader/full/berson-furan-maleic-anhydride-1953-ja01103a060 2/6
1‘722
JEROME
> I .
ERSON
ND RONALD
WIDLER
Vol. 75
thermodynamic stabilities are maintained. How-
ever, the literature pertaining to the additions o€
furan to maleic acid and maleic anhydride affords
no such detailed clarification of the factors controll-
ing the stereochemistry of the product. The
present communication reports results which place
the furan-maleic acid addition in the customary
energetic pattern.
Furan was observed to dissolve slowly when
agitated for several days with an aqueous solution
of maleic acid.8 Although the expected adduct
(1)
was not isolated from the reaction mixture,
bromination of the aqueous solution afforded a
nearly quantitative yield (“crude”) of the bromo-
lactonic acid IIa . By contrast, the reaction of
eOOH
CO
I
I
a,
R
=
H ;
I1 b, R = CH,s
furan with maleic anhydride in ether3 produced an
excellent yield of an adduct which was subsequently
shown‘o to be exo-cis-3,6-endoxo-A4-tetrahydro-
phthalic anhydride (I1I) .l1 In an att empt to
I11
establish the origin of this unique stereochemical
result, we have reinvestigated the furan-maleic
acid addition. Contrary to the previous implica-
tionla the reaction is not stereochemically homo-
geneous.
Analysis of the Reaction Mixtures.-The ease
with which both the endo- and exo-adducts of
furan and maleic acid disso~iate~~’~laces severe
limitations upon at tempts to isolate the substances
themselves. Accordingly, we have sought to
isolate them as stable derivatives. Since satura-
tion of the olefinic bond of the adducts renders
them incapable of diene retrogression,
our
first
analytical efforts involved attempts to fractionate
the crude product resulting from direct bromination
of the diene addition reaction mixture. The initial
crystalline material which precipitated upon
brom-
ination consisted largely of crude endo-bromo-
lactonic acid (IIa ). The mother liquors deposited
a mixture of
IIa
and an exo-dibromoacid (IVa)
which was identical with the substance previously
described by Woodward and BaerIo in their elegant
investigation of the bromination products of 111
(The rearranged exo-bromolactonic acid (Va) o
8) 0
Diels and
K.
Alder,
Ann . 490, 243 (1931).
(9)
0
iels and K. Alder, Be?.
62, 557
(1929).
(10) R.
B.
Woodward and H. Baer,
TEIS
OWRNAL, TO
1161
(1948).
(11)
Th e difference in the steteoch emicai courses
of
these two add -
tions has been attributed to a solvent dependence.#~.”
(12)
R.
. Elderfield and
T.
N. Dodd,
J r . ,
in “Heterocyclic Com-
pounds,’’ edited by R. C. Elderfield, John Wiley an d
Sons,
ne. , N e w
York,
N. Y. 950, p . l3Q.
0
11’
a,
R
=
H ;
I
b, R = CH; 1- , R
= CHI
V
a,
R
=
H;
V I
was not isolated in this procedure.) Separation
of IIa and IVa was facilitated by the discovery
that the sodium salt of IVa is virtually insoluble in
water. The exo-stereochemistry of IVa was con-
firmed by its independent synthesis from 111
Bromination of I11 in methylene chloride produced
the dibromo anhydride VI, hydrolysis of which
afforded IVa. VI was regenerated from IVa
by treatment with acetic anhydride or by pyrolysis.
Although the stereochemical heterogeneity
of
the
diene addition was thus unequivocally established,
this fractionation procedure was found inconven-
ient for the determination of the relative propor-
tions of t he isomers, since purification of the first
crop of crude IIa was achieved only with severe
losses
of
material.
A successful fractionation was finally based upon
the observation that addition
of
slightly less than
one equivalent of sodium bicarbonate to the re-
action mixture before bromination caused the pref-
erential precipitation of a considerable propor-
tion of the exo-adduct (VII) as well as unreacted
maleic acid in the form of a mixture (A) of sodium
salts. Hydrogenation of the aqueous mother
liquors (B) from the salt precipitation afforded the
satura ted endo-adduct (VIIIa) identified as the
corresponding anhydride (VIIIb) in small yield.
Hydrogenation of (A) yielded succinic acid but no
saturated exo-adduct could be isolated.
However, bromination of (A) produced virtually
pure exo-bromolactonic acid (Va). I t is of interest
that, in contrast to the procedure
(vide supra)
employing direct’ bromination of the reaction
mixture, bromination
of
the sodium salts (A)
produced little’or no dibromoacid (IVa) I 3 Brom-
ination of t he mother liquors B) produced mainly
IIa admixed with small quantities
of
Va.
Puri-
fication of IIa was accomplished by fractionation
from 90y0 methanol. The course of purification
was readily followed, since the presence of even a
small quantity of Va in solid samples of IIa is
sufficient to cause darkening upon fusion. Accord-
ingly, fractionation was continued until the ma-
terial melted to a colorless liquid. An additional
quanti ty of endo-derived material was obtained by
esterifying the combined tail fractions and frac-
tionating the mixed methyl esters (IIb and Vb).
(13)
In
most
of the runs, IVa could not be isolated via
its
sodium
salt)
Only in
runs
in which the embadduct concentration was high did
IVa
appear at all, and then only in quantities never exceeding 0 005
mole
per
mole of
Va.
8/10/2019 Berson Furan Maleic Anhydride 1953 Ja01103a060
http://slidepdf.com/reader/full/berson-furan-maleic-anhydride-1953-ja01103a060 3/6
April 5 1953
THE
STEREOCHEMISTRY
O F
THE
FURAN-MALEICCID
REACTION
1723
Although IIb was obtained in pure form, a small
tail fraction
(C)
consisting of an inseparable mix-
ture of esters always remained. This mixture
remained almost constant in weight and melting
point throughout several runs regardless of the
relative amounts of exo- and endo-derived material
obtained in the first part of the separation scheme.14
Results
and
Discussion.-The gravimetric values
for product composition as a function of time are
given in Table
I
and the results are shown graphi-
cally
in
Fig. 1. The somewhat greater scattering of
the points in the initial portion of t he curve (up to
72
hours) is presumably attributable to the less
efficient thermostating used during the period when
the reaction mixture is heterogeneous (see the
second paragraph of the Experimental section).
The total yields
of
isolated products as a function
of time are plotted in Fig. 2 . Although some ir-
reversible destruction
of
furan (by hydrolysis in the
acidic medium and self-condensation
of
the hy-
drolysis products) undoubtedly occurs as evidenced
by the darkening of the reaction mixtures and by
the observed presence of carbonyl bodies, this
process must be slow compared to diene additions
and retrogressions of t he adducts, since Fig.
2
shows no downward drift in total yield over a
period greater than one-third of the total reaction
time examined.
TABLE
PRODUCT
OMPOSITION
N THE FURAN-MALEICCID RE-
ACTION
Time, C X 0 , b
cndo,c
Tota l yield,d
hours
mole Va
mole IIa and IIb
23
32
48
55
72
91
105
118
165
213
285
. . .
. . .
0.018
. .021
.025
.038
.051
,061
,076
,092
,101
0.061
.066
.068
.075
.080
.079
.078
.072
.067
,054
.043
30
32
42
47
51
57
63
65
70
71
70
a All runs were a t 27.5 0.7 to 72 hours, at 27.5 f 0.1
afterwards. Values are for single determinations. From
preliminary duplicate runs, the reproducibility of a given
value in moles was estimated
at
f O O O 1 mole; reagents:
furan, 0.205 mole; maleic acid, 0.306 mole. The weight
of IVa was neglected. C T h eweight
of
IIa was corrected
for hydrate water. Pur e products only. The methyl
ester tail fraction (C) was neglected.
'The variation of product composition with time
(Fig. 1) unambiguously establishes the furan-
maleic acid reaction in the typical kinetic-energetic
pat tern. endo-Addition is the rate-favored proc-
ess but the endo-adduct is thermodynamically
unstable with respect to the exo-adduct which
gradually accumulates a t it s expense.
Two kinetically discrete sets of processes have
been proposed to account for the isomerization of
14)This may be due, at least in part, to th e apparent ability of IIb
and
Vb
to form a crystal compound. A melting point 8s. composition
diagram of synt hetic mixtures
of
IIb
and
Vb
showed a maximum in the
region of
ca.
40Y0
f
Vb. However, since
Vb
decomposes upon fusion,
the observed melting poinls are somewhat dependent upon the rate of
heating and the curve has not been reproduced here.
0.09
0.07
0.05
ai
-
0.03
0.01
0 50 100 150 200 250 300
Time, hours.
Fig. 1.-Variation
of
product composition with time in the
furan-maleic acid reaction.
r
1
$ 3 0 p.
G I
10
0 100 200 300
Time, hours.
Fig. 2.-Variation of tota l yield with time.
endo-diene adducts to exo-isomers. The first
of
these, proposed3,l5 for the interconversion of the
stereoisomeric adducts in the fulvene series, con-
siders tha t the isomerization occurs by dissociation
of the endo-adduct to the addends which then re-
combine at a slower rate to give the relatively
stable exo-adduct. The process is summarized in
e'quation (1). CraigI6 has proposed an alternate
mechanism to account for certain phenomena in the
exo-adduct addends endo-adduct
(1)
interconversion of the cyclopentadiene-maleic an-
hydride adducts (IXa and b). This author sug-
gests that the endo-adduct (IXa) is isomerized to
the exo-adduct (IXb) directly through some non-
isolable intermediate without prior dissociation
to the addends. While certain details of the in-
timate mechanism
of
this process suggested by
Craig appear implausible
to
us
on
energetic grounds,
the general nature of t he scheme, summarized in
equation
(2)
is not refuted by any da ta now avail-
able in the literature.
If
both endo- and exo-
addends endo-adduct
_
exo-adduct (2)
adducts are formed directly from addends in simul-
taneous, competing reactions (equation
l ,
the
( 1 5 ) K. lder and W. Trimborn,
A X E . ,
66,
58
(1950).
(16) D. Craig, THISJOURNAL 1 8 ,
4889
(1951).
8/10/2019 Berson Furan Maleic Anhydride 1953 Ja01103a060
http://slidepdf.com/reader/full/berson-furan-maleic-anhydride-1953-ja01103a060 4/6
1'724
JEROME *I. ~ E K S O NNI)
RONALDWIDLEK Vol.
75
slopes (dce,d,/dt and dcexodt) of the adduct formed
BS.
time plots for the two isomers should have
maximum values when the addend concentration
is highest, Le., a t the beginning of the reaction.
This follows from the rate expression (assuming
second-order kinetics), dc/dt = k [furan] maleic
acid].
If
the situation is represented by equation
i n ,
however, dcen(io t will have a niaximum value
at the onset of reaction, bu t dcEno t will gradually
increase and will reach i ts maximum value a t a
time corresponding to or following the time at
which the concentration of endo-adduct is at a
maximum. Inspection
oi
Fig. 1 reveals that
dcexo,'dtgradually increases with time and that an
inflection occurs in the general neighborhood
of
the
maximum endo-adduct concentration. While this
result seems to provide support for a set of proc-
esses governed by equation 2 ) , it must be em-
phasized th at the non-availability of the endo-
adduct per selohas prevented a test of our analytical
scheme upon synthetic mixtures of known com-
position. Therefore, the possibility exists that
serious preferential manipulative losses of exo-
material a t low exo-adduct concentrations are to
blame for the apparent shape of the curve.
decision between the mechanisms represented by
equations
( I )
and
(2)
cannot, therefore, be made
on the basis of the present da ta.
In order to make an approximate evaluation of
the equilibrium constant for the reaction endo
exo in the present case, we have allowed the pure
exo-adduct (VII) to stand in solution a t room
temperature for varying lengths of time.
-4n
appropriate quantity of excess maleic acid was
added to the reaction mixtures
so
that the analy-
tical conditions would simulate those existing in the
diene addition runs, and the reaction mixtures
were analyzed by the scheme described above.
The gravimetric values are indicated
in
Table I1
and plotted in Fig. 3. The recovery of only 88%
of exo-material even when
no
endo-product could
be isolated probably represents the limit of accuracy
of the method a t high exo-adduct concentrations,
although a part of the unrecovered 12y0of material
may be in the form of addends generated by retro-
gression. The small decrease in tota l recovery
from
261
to 1304 hours probably is attributable to
some irreversible destruction of furan. (The
reaction mixture becomes quite dark after pro-
longed standing.) IYhile the final figures of Table
018011 ' '
I
O 0 G O
END0
0.020
0
500
1000
Time, hours.
Fig. 3.-Equilibration
of V I I .
I1
(1304 hour run) probably
do
not represent the
true equilibrium ratio, it is clear that the experi-
mental equilibrium value lies somewhere between
these figures and those of the 285 hour run of Table
I. That the concentration changes after
1304
hours of equilibration are very slow is indicated
by the shallow slope of the curves of Fig.
3.
The
failure to achieve quantitative material balance in
the equilibration
runs
prevents, of course, the
accurate estimation of t he equilibrium constant for
the isomerization, K i = [exo]/ [endo]. However, a
maximum value of K i is obtainable by assuming
the unrecovered material a t 1304 hours t o be of
exo-origin. This gives a value for
i
of
0.181/
0.024 =
7.55, a figure which can only be diminished
by a closer approach to equilibrium than was
achieved after 1304 hours. From the relationship
A F i =
-RT
In K i , the free energy change for the
interconversion in water at
28
cannot be larger
than ca.
.2
kcal./mole.
TABLE
1
PHTHALIC
ACID ( V U )
IN
AQUEOUS
MALEIC
CID AT
28
h o u r s
mole
Va
mole IIa and IIb 70
20 0.180
. . . 88
261 . 1 3 5 0.013 7 2
1304
.114 ,024
67
Reagents: em-adduct
(VII),
0.205 mole (prepared from
0.205 mole of 111); maleic acid, 0.100 mole (prepared from
0.100
mole
of
maleic anhydride); water
75
cc.
On the basis of available i n f o r m a t i ~ n ~ - ~ ' ~ ~t seems
likely that the energy differences in previously
observed pairs of endo-exo isomers are not very
much greater than the maximum value of
1.2
kcal./mole observed in the present case. I t is
striking tha t, despite the small differences in energy,
the thermodynamic stability order (ex0 > endo)
persists in all the cases studied
so
far. The reasons
for this are not immediately apparent.
A
qualita-
tive estimate of the stability relationship in the
case of the completely saturated pair of isomers
derived from dicyclopentadiene (Xa and Xb) may
be made on the basis of the non-bonded interaction
effects believed18-21 to operate in alicyclic and
paraffinic hydrocarbon derivatives. The boat
form of cyclohexane (X I) may be considered as the
simple prototype of Xa and Xb. In XI , by the
conformational criteria developed18 for the cases
of methylcyclohexane and the dimethylcyclohex-
anes, the energy of interaction
of
an equatorial
(e)
methyl group
a t CB
or C3with a polar
(p)
methyl
group at C1or Cd, respectively, will be equivalent
to tha t of a $-methyl group at Cz or
CS
with a
p-hydrogen at Cg or
C,,
respectively. I t seems
reasonable to suppose that the same relationship
would hold for the methylene-methylene interac-
tion as compared to methylene-hydrogen with the
exception that the actual values of the energy for
each such interaction will be slightly smaller than
EQUILIBRATIOK O F
exo-Gis-3,6-ENDOXO-~. -rETRAHYDRo-
Time,
exo, ewdo,
Tota l yield,
(17)
K.
Alder and G. Stein, Be?., 67, 613
(1934).
(18) C . W. Beckett, K. S. Pitzer and R. Spitzer, THISOURNAL 69,
(IS)
C.
W. Beckett,
N. K.
Freeman and
K.
S.
Pitzer,
ibid.
70,
20)
0.
Hassel and
B.
Ottar, Ac t a Chetn. S c a m . , 1 9 2 9 (1947).
2 1 ) D. H . K.Barton,
Ewpe v i e i r t i n
6 , 316 (1050).
2488 (1947).
4227 (1948).
8/10/2019 Berson Furan Maleic Anhydride 1953 Ja01103a060
http://slidepdf.com/reader/full/berson-furan-maleic-anhydride-1953-ja01103a060 5/6
April 5, 1953
THE
TEREOCHEMISTRY
OF
THE FURAN-MALEICCIDREACTION
1725
a
the value of 0.9 kcal. suggestedI8 for the methyl-
methyl and methyl-hydrogen cases. However, in
Xa and Xb the substituents a t CI and Cd no longer
have the true polar conformation but are displaced
to a distorted conformation which we have desig-
nated p'. A group in the p'-conformation
Cy)
is further away from an e-substituent at
CZ
or C3
than is the case with a group in the p-conformation
and therefore the non-bonded interaction will be
less.22 On the other hand, interaction effects
between p-substituents at Cz or
C3
with p-hydro-
gens at Ce or C6 espectively, would be expected
to retain their normal values. Assuming (i)
that the distortion introduced by the endomethyl-
ene bridge
has
the maximum conceivable
effect
.and completely eliminates two sources of non-
bonded interaction between
exo-C2
and Ct sub-
stituents and the bridge methylene group, (ii)
that each interaction contributes a maximum of
0.9 kcal. of Pitzer-strain, and (iii) tha t the only
source of difference in heat content between Xa
and Xb is ('Pitzer-strain,'' then we may estimate
the maximum difference in heat content for this
pair of endo-exo isomers to be not greater than 1.8
kcal./mole.23
While similar conformational considerations
apply to diene adducts themselves, the argument
requires modification with respect to several new
factors which are not yet capable
of
easy evaluation.
Thus, in a diene adduct such as IXa , the ?r-orbitals
of the olefinic bond and of the carbonyl groups are
directed toward each other. Whether this will
result in a mutual net attraction (through disper-
sion forces, for example) or repulsion of the groups
seems difficult to assess. Further, the occurrence
of strong dipole effects arising from the presence of
polar substituents, solvation effects and intra-
molecular hydrogen-bonding (particularly in a
case such as the furan-maleic acid adducts) may
take energetic contributions of far greater magni-
tude than the non-bonded interactions. It seems
entirely conceivable that, under the proper cir-
cumstances, these and other effects may supple-
ment or oppose each other
so
that the energy
differences between endo- and exo-isomers
wll
be
greater than have previously been observed or,
conversely, so that there may be a reversal of the
stabili ty order.
(22) The e'-hydrogens are correspondingly closer to e-sub stitu ents
at
C2
or Cj.
(23)
The argument applies strictly only to heat content. The com-
parable rigidi ty and symmetr y of a pair
of
endo-em-isomers, however,
justifies the omission of entropy considerations (cf. ref.
3)
and we there-
fore have assumed in the sequel that free energy values (derivable from
equilibrium data) will be affected to a similar exten t.
This interaction is neglected.
XI
Experimentalz4
Reaction of
Furan
with Maleic Acid.-The reaction ves-
sels were stout 150-cc. ground-glass stoppered bottles of uni-
form dimensions. Maleic acid was prepared by dissolving
the necessary quantity
(30.0
g., 0.306 mole)
of
freshly re-
crystallized maleic anhydride in
75
cc. of water in each re-
action vessel. The contents were chilled to
-5'
and to
each bottle was added
15.0
cc.
(0 .205
mole) of redistilled
furan (du Pont) . The bottles were sealed by heavily tap-
ing the stoppers and dipping the heads of th e vessels into
molten paraffin.
The bottles were shaken in a constant temperature room
a t
27.5
=k
0.7'
on a mechanical shaker. The furan slowly
dissolved and apparent homogeneity was achieved after
about
72
hours, whereupcp the bottles were transferred to a
thermostat a t
27.5
i .1 .
em-Bromolactonic Acid (Va) -Each reaction mixture
was worked up as follows: Th e bottle was chilled to O' ,
opened and the contents diluted to
125
cc. with water.
The solution was treated a t 0 with
22.6
g. of solid sodium bi-
carbonate (added in portions during 10 minutes). The
colorless precipitate ( A ) was collected a t the pump an d
washed with two 25-cc. port ions of ice-water. The filtrate
(B) was preserved for the subsequent isolation of IIa. The
salt ( A ) was suspended in
125
cc. of water, the mixture
chilled to
5 ,
and bromine was added until
a
faint yellow
color persisted. The reaction mixture was made strongly
acidic with concentrated hydrochloric acid and treated with
solid sodium bisulfite to decompose the excess bromine.
The product was collected
at
th e pump, redissolved in 10%
potassium carbonate, reprecipitated with concentrated hy-
drochloric acid, collected and dried n vacuo to constant
weight over calcium chloride. This material melted sharply
a t
158'
with violent decomposition when it was slowiy
heated
(0 .5
per minute) after insertion into a bath a t
140
,
[reportedlo m.p.
153
(dec. )]. It s infrared spectrum was
identical with th at of an authentic specimen prepared by the
method of Woodward and Baer.10 There was no change in
melting point or infrared spectrum upon recrystallization
from water. In the
runs
a t longer reaction times, trace
amounts of the dibromoacid (IV a) were isolated from the
mother liquors by the sodium salt precipitation technique
(vide infra). For example, in a run in which
28.5
g. of Va
was obtained,
0.25
g. of IVa was isolated.
endo-Bromolactonic Acid (Ira).-The filtrate (B) from the
bicarbonate precipitation was treated with
10
g. of solid
sodium bicarbonate, chilled to 5 , swirled and treated with
bromine unti l the color of excess bromine persisted. The
clear solution was slowly acidified with concentrated hydro-
chloric acid, the excess bromine decomposed with solid
so-
dium bisulfite, and th e mixture cooled to 0'. The precipi-
ta te of crude endo-bromolactonic acid ( II a) was collected
a t the pump and dried in vacuo over calcium chloride for
48
hours. This material melted at 200-204' with pre-darken-
ing
and extensive decomposition. Recrystallization from
90%
methanol, and then from water, gave glittering, bold
staves, melting at
207-209'
to a colorless liquid, reported*
m.p.
205'.
The melting point was unchanged by further
recrystallization. This material was dried to constant
weight over calcium chloride in vacuo.
(24) Meltin g point s are corrected. Th e infrared spectra were deter-
mined in Nujol mulls with a Baird Associates Recording spectrometer.
We are grateful to Dr.
H.
L. Holmes and his associates
of
Riker Labora-
tories, Inc.,
for
the measurement of the spectra. The microanalyses
were performed by Mr. Joseph Pirie.
8/10/2019 Berson Furan Maleic Anhydride 1953 Ja01103a060
http://slidepdf.com/reader/full/berson-furan-maleic-anhydride-1953-ja01103a060 6/6
1726 JEROME
-1.
ERSOK
ND
R O N A L DWIDLER Vol. 7 5
endo-Bromolactonic Ester (IIb).-The methanol mother
liquor from the purification of II a was evaporated in a cur-
rent of dry air and the solid residue dried in z~acuo ver cal-
cium chloride and potassium hydroxide. The dried prod-
uct was taken up in 100 cc. of absolute methanol, decolor-
ized with charcoal, filtered, treated with 2 cc. of 30Yo fuming
sulfuric acid and heated at reflux for two to three hours.
The mixture was cooled to
O ,
the solid ester collected at the
pump and recrystallized from absolute methanol to constant
melting point . The pure ester was obtained as glistening,
micaceous platelets, m.p. 176-177 , reporteds m.p. 175'.
Evaporation of t he methanolic mother liquors left a small
residue melting a t 150-155 . Attempts to obtain pure
material by fractional crystallization from a variety of sol-
vents were generally unsuccessful.
Direct Bromination Procedure.-If the diene addit ion re-
action mixture was brominated directly, without prior addi-
tion of sodium bicarbonate, crude II a separated from
solution as a white precipitate, melting over a range from
ca. 180-200 . Recrystallization from water produced pure
material
but was attended by severe losses. Thus , in a
typical run , from 32 g. of crude precipitate was obtained
9.5 g.
of pure Ila.
The bromination mother liquors were cooled to 0 for
21 hours and the second crop of crystals collected. These
usually melted a t 141-148 with decomposition. The solid
was triturated with 10% sodium bicarbonate and filtered
Acidification of the basic mother liquor produced a sinall
additional quan tity of I Ia . The bicarbonate-insoluble
filter cake was warmed with concentrated hydrochloric acid
and the resultant solution, upon being cooled, deposited
shining, transparent laths of the dibromoacid (IV a) hydrate,
m. p. 163-164'. (The m.p. seems to depend on the rate of
heating; occasionally
a
m.p. of 159-160 was observed;
reportedlo m.p. 155O.) A
mixed m.p. with material pre-
pared by the method of Woodward and Baer'o was not de-
pressed. No Va was isolated. Recrystallization of IVa from
a minimum quantity of water failed to change the melting
point . The analytical sample was dried at room tempera-
ture for eight hours a t 3 mm. over phosphorus pentoxide.
Anal.
Calcd. for C8HsOiBr2.H20:C, 26.54;
H,
2.98;
Br, 44.15. Found: C, 26.92;
H,
2.96; Br, 43.89.
After drying for four hours at 100' and 3 mm. over P205
the crystals became opaque, m.p. 163-164 .
Anal. Calcd. for CgH80jRr2:C, 27.93; H, 2.34. Found:
C, 27.79; H, 2.46.
The acid (IVa) dissolved readily in concentrated aqueous
ammonia. When this solution was treated with saturated
sodium chloride solution, the sodium salt of IVa precipi-
tated immediately. Treatment of the solid salt with hy-
drochloric acid regenerated IVa.
The dimethyl ester
IVb)
melted at 114.~-l15.5°,e-
porkdl0 m.p. 115.9-116.3 .
Mbromoanhydride (VI). A.-To a solution of
16.8 g.
of th e furan-maleic anhvdride adduct ( I I I ? in 400 cc. of
dry, freshly distilled tneihylene chloride was added 5.2 cc.
of bromine in one portion. The reaction mixture was kept
a t room temperature for four hours, after which time the
color of the bromine had disappeared. The crystalline pre-
cipitate (10.0 g.) , m.p. was filtered
off
and the mother
liquor evaporated to dryness. The residue weighed 22.1 g.
and melted at 187-160 . Recrystallization from chloro-
form-acetone 2fforded 18.6 g. of silky, clustered blade:,
m.p. 159-160 . The analytical sample, m.p. 162-163 ,
was recrvstallized from carbon tetrachloride-acetone.
Anal. CaIcd. for C8H&laBr2: C, 29.48; H , 1.86. Found:
C ,
29.50; H, 1.96.
Hydrolysis of the anhydride in boiling water produced
the dibromoacid (IVa), m.p. 163-164 alone or mixed with
isample prepared as above or by the method of itroodward
and Baer.'O
B.-A mixture of 3.5 g. of IVa in 15cc. of acetic anhydride
was heated a t reflux for one hour. The reaction mixture
was diluted with 20 cc. of chloroform, evaporated to 15 cc.
volume,
treated with an additional 20 cc.
of
chloroform,
filtered hot and allowed to cool. The product (0.9 9.) had
m.p. 161.5-162.5' alone or mixed \Vith a sample of 1-1 pre-
pared by method A .
C.--Two grams of IVa was heated above its melting point
for ten minutes. The pyrolysis residue was sublimed a t a
Its nature is now being
______-
( 2 5 ) This substance is isonieric
with
V I .
investigated.
pressure of 1 mni. and a bath temperatur: of 165
to
yield
1.7 g. of
V I ,
m.p. and mixed m.p. 159-161 .
Hydrogenation of Diene Addition Reaction Mixtures.-
The reaction mixture w-as subjected to preliminary fractiona-
tion with sodium bicarbonate (vide supra) . The mother
liquor ( B ) was hydrogenated over 0.5 g. of palladized ba-
rium sulfate (10%) at an initial pressure of 35 lb./sq. in.
After about 0.03 mole of gas had been consumed, 0.1 g. of
platinum oxide was added and hydrogenation was continued
until t he total consumption of gas amounted to 0.144 mole
whereupon the reaction ceased spontaneously. The solu-
tion was filtered, acidified with concentrated hydrochloric
acid and cooled. The precipitated solid was heated above
its melting point for ten minutes and recrystallized from
acetic anhydride-Skellysolve B to yiekd 2.5 g.of the endo-
cis-anhydride (VIIIb), m.p. 138-160 , reported26J7 m.p.
Hydrogenation of the sodium salt precipitate (A) from the
bicarbonate fractionation in aqueous suspension over plati-
num oxide yielded succinic acid as the only isolable product.
Epimerization of the endo-Bromolactonic Acid ( IIa)
--
The procedure of Diels and Alders for epimerizing IIa was
found to be inconvenient.
We
have prepared the epimeric
acid (XII) by a new method. A mixture of 10 g. of I Ia
158-159 .
Y
XI1 a;
R = H;
XI1 b;
R =
CHI
and
5 J
cc. of acetic anhydride was heated a t reflux for
15
minutes. The clear solution was then concentrated by dis-
tillation until the vapor temperature reached 136 . Acetic'
anhydride was added to restore the volume to 50 cc. The
solution was then heated a t reflux an additional 30 minutes,
and the concentration and dilution procedure repeated.
After an additional four-hour period of boiling, the solution
was concentrated to half its volume, treated with 75 cc.
of
water to decompose the acetic anhydride and allowed to
cool. The epimerized acid (XI Ia ) separated as 7.2 g. of
white needles, m.p. 236-238', reported* m.p. 23;-232'.
The methyl ester (XIIb) had m.p. 170-171
,
reported8
m.p. 167-168'.
Application of this epimerization procedure to th e methyl
ester of the endo-bromolactonic acid (II b) resulted in re-
covery of 88 of unchanged starting material.
endo-Dilactone (XIII)-The procedure used for prepara-
tion of this substance was based upon the one used by Alder
and Stein for the preparation of the analogous dilactone
(XIV) from the endo-adduct of cyclopentadiene with maleic
anhydride. To a solution of 3.9 g. of potassium hydroxide
in 50 cc. of methanol was added 5.2 g. of I Ia . The mixture
was heated to reflux and
a
colorless precipitate appeared.
After two hours, water was added to dissolve the precipi-
tated salt, and the solution was neutralized to congo red
with concentrated hydrochloric acid. The solution was
evaporated to dryness and the residue heated on the steani-
bath with 50 cc. of acetic anhydride. The reaction mixture
was filtered from the insoluble inorganic residue and allowed
to cool. The dilactone (XIII ) crystallized as
0.75
g. of
stout, sharply defined rhombohedra, m.p: 261.5-265.5'.
Recrystallization from acetone-methanol raised the m.p. to
269.5'. The substance is insoluble in cold 10% potassium
carbonate.
And .
Calcd. for C8H8Ob: C, 52.75; H , 3.32. Found:
C ,
52.49;
H,
3.28.
Los
ANCELES. CALIFORNIA
(26) K. Alder and K. H. Backendorf, A n n .
535,
113 (1938).
(27) Cf.
eference 10
for
the correct configurational assignment.
( 2 8 ) K.
Alder and G. Stein,
An n . 514, (1934).